Dear All,
I have data on organ donation, which I think is the universe of organ donors and recipients in the US. I am studying the difference between males and females, and unfortunately balance tests failed in the sense that many covariates are significantly correlated with gender. I understand that there is a debate about the value and legitimacy of balance tests, but I am not interested in the debate. Rather, assuming that balance tests are an important part of identification strategy, my question is if balance tests are irrelevant or still relevant to my data, given that it is the universe, not a sample, random or not.
Thanks,
Kangoh
I have data on organ donation, which I think is the universe of organ donors and recipients in the US. I am studying the difference between males and females, and unfortunately balance tests failed in the sense that many covariates are significantly correlated with gender. I understand that there is a debate about the value and legitimacy of balance tests, but I am not interested in the debate. Rather, assuming that balance tests are an important part of identification strategy, my question is if balance tests are irrelevant or still relevant to my data, given that it is the universe, not a sample, random or not.
Thanks,
Kangoh
Comment